Abstract Structurally dissimilar aggregates (strains) of the amyloid beta peptide (A?) in Alzheimer's Disease (AD) can potentially explain differences seen in the progression and severity of the disease. Fibrils formed by synthetic A? in vitro and A? fibrils seeded from patient brain extracts led to a variety of A? strains. Previous research with AD brain seeded material led to A? structures that varied with patients and with the stage of disease, suggesting that specific A? strains not only could affect the progression of AD but also potential treatment. However, a bias using patient seeded synthetic A? is that seeding might select for those A? strains with the highest seeding potential masking other strains that could be important in the disease. Our long-term goal is to understand the basis of strain variation for several pathological proteins important in neurodegenerative disease such as tau, ?-synuclein, huntingtin, and A?. The objective of this application is to determine the in vivo-generated structures of A? strains found within and between individual amyloid mouse models, which will answer the following questions: 1) Are A? plaques found in individual mice composed predominantly of one strain or mixtures of strains? Do strains depend on gender, brain region, or mouse model examined? How are strains impacted by seeding mice with fibrils from human brains? 2) Are seeding experiments capturing the structural variety found in AD brains or are they biased towards the most seeding competent species? 3) what are the structural differences between these strains? We will address these questions in the following 3 specific aims: In Aim 1, we will directly detect the A? strain variety and distribution in mouse models of amyloid pathology. This will be accomplished by measuring solid-state NMR spectra on brain extracts purified from 15N labeled APPKINL-F, APPKINL-G-F, and 5XFAD mice. A subset of 5XFAD mice will be seeded with AD patient brain extract. In Aim 2, we will determine the seeding potential of A? strains from amyloid mouse models. We will seed recombinant A? with brain extract from amyloid pathology mouse models, measure the seeding kinetics of different strains, and compare their NMR spectra to those of the original mouse brain extract and those of A? seeded from human AD brain extract. In Aim 3 we will determine the structures of a basis set of A? strains from amyloid mouse models. We will use an innovative solid-state NMR and EPR approach to determine high- resolution structures of A? that capture short and long-range order details. The expected outcome of these aims is that we will pioneer NMR spectroscopy on vivo-generated A? aggregates. We will map the distribution of A? strains throughout the brain, and determine the dependence of strains on brain region, gender, age, and mouse model. We will correlate the NMR structures with brain pathology by histology and biochemistry. We will develop a combined EPR and solid-state NMR approach for fibril structure determination. These outcomes will enable us to design A? strain dependent diagnostics and treatment for AD.